< back

Week 6: Molding and Casting

Bicycle Paper Weight

October 25, 2015

This week's task was molding and casting. I felt a lot of different constraints this week (most notably the time constraints of impending Member's Week and the space constraints of a 3" x 3.5" wax block) so I don't feel I really got to do something as creative or interesting as I would have liked. Still, my work this week was a good opportunity to get acquainted with the process of molding and casting. For my mold, I toyed with the idea of making some sort of jewelry or trying to cast the Golden Gate Bridge, but I couldn't settle on a jewelry design I liked, and the Golden Gate Bridge seemed to have overhang on too many axes to be able to mold it. Hopefully I'll be able to return to some of these ideas in the future! For this week, though, I just decided to make a simple bicycle "paper weight" or figure. To make my model, I imported a bicycle vector file into Rhino, and extruded it to make the positive form for milling. I realize that this design is largely 2.5 axis milling (instead of 3 axis milling) - that is another complexity I hope to return to in the future. I fit the bike model well within the size of my block, and also added registration marks, and a sprue hole and air vent. I exported the models as STLs to be opened in Partworks 3D.

In Partworks 3D, I imported my STLs and went through the steps we had discussed in training to prepare my toolpaths. I thankfully didn't have any issues with units this time around (I modeled in inches and was able to import my STLs in inches), but I still encountered a few complexities in the software. My biggest problem was that I initially couldn't get the software to leave "walls" around my mold. I eventually solved this problem by setting the "machining margins" to 0 on all sides. My model already had some built-in space between the edges of the bike and the edge of the plane it was lying on, so I think that's why I was able to set this value to 0. Otherwise, I used all standard values that had been discussed in training:

After exporting my toolpath, I fixtured my wax to the Desktop Shopbot by gluing the wax to a piece of OSB, then affixing the OSB to the machine bed with screws. I zeroed the Shopbot using the same method as I used with the big Shopbot. For the Desktop Shopbot, you don't need to manually start the spindle, so I just loaded my parts in the Shopbot software and sent them to the machine.


I flipped the wax over to make the second mold. Both pieces came out pretty nicely! The wax is a really nice material to mill. I had to use an Exacto knife to remove a few extra pieces of wax in places the mill was not able to reach. I'm a bit worried that the walls of the bike will be too thin for the drystone to reach or that the drystone won't flow throughout the whole model. I suppose I'll cross that bridge when I get there.

Setting up the molds was pretty simple. I used the standard silicone rubber, Oomoo, for my molds. Oomoo has two parts, A and B, that are mixed in a 1:1 ratio. You have about 15 minutes to work with the Oomoo before it starts to cure, which gave me ample time to carefully stir the parts individually, and then combined. It's important to stir/ pour the Oomoo slowly to avoid introducing bubbles. You can also "tap" the cup or your mold on the table when they're holding Oomoo to help get some bubbles out. I used cardboard to build up the walls around my milled design and poured my first mold, which I left to cure for about 2 hours.

I went back to check the mold after 2 hours, and unfortunately it still felt pretty gooey. Perhaps it didn't cure all the way because the molding and casting room is so cold. I went back to check in the morning and found the mold had cured completely. Satisfied, I poured the other mold and left it to cure overnight.

The second mold came out quite nicely as well. Unfortunately, as soon as I tried to put them together to cast the bicycle, I realized I had made a huge mistake! I made the molds equivalent instead of mirror images of each other. This meant that the two-part mold wouldn't fit together and I couldn't cast my bike in it. I hadn't thought much about mold orientation while modeling (obviously), but it is definitely something I will consider more carefully next time - alas! Instead of using the two-part mold, I cast half the bicycle in Drystone in one part of the mold. Using a one-part mold made it much easier to pour the Drystone and make sure it was distributed in all parts of the mold. Due to the simplicity of my model, the one part mold cast should look quite similar to what would have come out of a two-part mold - same shape, just a bit skinnier!

I de-molded the Drystone after about 3 hours. Unfortunately, the cast bicycle broke in a few places where the model was especially thin. I didn't realize how brittle this material is. It's also possible I used too much water when preparing the Drystone - I mixed it by feel and it was, if anything, a little watery. I also cast the bicycle in Hydrostone, but the cast fractured when I tried to remove it from the mold. Lessons from this molding/ casting session:

I also started to experiment with making molds from Lego bricks. I cast four Lego bricks in a block of Oomoo, using clay to make sprues and air vents. Once the Oomoo was set, I was able to cut the Oomoo into the four pieces and split those pieces in half to make individual molds for each Lego piece. These molds came out very nicely, preserving small details from the Lego bricks. I cast two bricks in Drystone. As with the bicycle mold, I was worried that the walls of the Lego pieces would be too thin to cast successfully.

The casting actually went much better than expected. The bottom of the bricks are beautiful, smooth, and don't have air bubbles. The tops were less successful...in most places, the Drystone wasn't able to get up into the mold for the "bumps" on top of the bricks. However, I was able to file off the Drystone that had set in the sprue to reveal a few "bumps", which allowed me to actually connect the bricks!


Design Files